DeepSeek, the Hangzhou-based AI lab known for its unusually efficient open-weight models, has released DSpark, an open-source framework that it says can accelerate large language model (LLM) inference — the process of actually running a trained model to answer queries — by up to 85%, according to a VentureBeat report published June 28, 2026.
The release continues DeepSeek’s pattern of publishing its internal efficiency tooling openly rather than keeping it proprietary, and lands at a moment when inference, not training, has become the dominant cost line for companies serving AI at scale.
Executive Summary
The announcement is straightforward on its face: DSpark is an inference framework, it is open source, and the headline claim is a speedup of “up to 85%.” What makes it noteworthy is who is making the claim. DeepSeek built its reputation on doing more with less — its earlier model releases were credited with achieving frontier-class results at a fraction of the compute budgets reported by Western rivals — so an efficiency claim from this lab gets taken more seriously than the average vendor benchmark.
If the speedup holds up under independent testing, the implications run well beyond one company’s software stack. Inference speed translates almost directly into serving cost: a model that answers queries faster on the same hardware serves more users per GPU, which means fewer GPUs, less power, and less data center capacity per unit of AI demand. Because DSpark is open source, any operator — hyperscaler, neocloud, or enterprise running models in-house — can in principle adopt it without a licensing negotiation.
The important caveat is that “up to 85%” is a ceiling, not an average, and the report available at publication does not detail the workloads, models, or hardware behind the number. That distinction should shape how buyers and investors read the news.
Inference Is Where the Money Now Goes
For the first years of the generative AI boom, the eye-watering costs were in training — the one-time process of teaching a model from massive datasets. That has flipped. Once hundreds of millions of people are querying models daily, the recurring cost of inference dwarfs the one-time cost of training, and it scales with every new user and every longer conversation. This is why the industry’s optimization energy has shifted to serving: techniques with names like speculative decoding, quantization, and KV-cache management all exist to squeeze more answers out of each GPU-hour.
An 85% speedup, if achieved on realistic workloads, is not an incremental gain in this context. Serving capacity is the binding constraint for many AI providers, and GPUs remain supply-limited and expensive. Software that meaningfully raises throughput per chip is functionally equivalent to manufacturing more chips — without the fab, the lead time, or the export-control exposure that hardware carries.
DeepSeek’s Open-Source Playbook, Continued
DeepSeek has a track record here. The lab, spun out of the Chinese quantitative hedge fund High-Flyer, shook global markets in early 2025 when its R1 reasoning model demonstrated that frontier-adjacent capability did not require frontier-scale budgets. It followed up by open-sourcing chunks of its internal infrastructure code — low-level GPU kernels and communication libraries — rather than treating them as trade secrets. DSpark fits that pattern: release the tooling, let the ecosystem adopt it, and compete on the pace of research rather than on locked-down software.
The strategic logic is worth spelling out. Open-sourcing inference tooling commoditizes the serving layer, which pressures companies whose business model depends on proprietary serving efficiency, while costing DeepSeek little — its own advantage lies upstream, in model quality and training efficiency. It also builds developer mindshare globally at a time when Chinese AI labs face restricted access to top-end accelerators, making software efficiency a competitive necessity as much as a virtue.
What Cheaper Inference Means for Infrastructure Operators
A natural first read is that faster inference is bearish for GPU and data center demand: if each chip does 85% more work, you need fewer chips and fewer megawatts. History suggests the opposite usually happens. Efficiency gains in computing have repeatedly triggered what economists call the Jevons paradox — when something gets cheaper, consumption expands enough to more than offset the savings. Cheaper inference makes previously uneconomic AI applications viable: always-on agents, AI in low-margin consumer products, long-context document processing at scale.
For data center operators and connectivity providers, the more defensible conclusion is that efficiency software shifts demand rather than shrinking it. Lower serving costs favor deployment breadth — more applications, more regions, more inference happening closer to users — which tends to benefit distributed capacity and network infrastructure even if it moderates the growth rate of any single mega-campus. Operators planning around raw GPU scarcity should note that the scarcity premium softens every time the software stack gets meaningfully better.
Reading an ‘Up To’ Claim Responsibly
The 85% figure deserves the same scrutiny any vendor benchmark gets, and the fact that DSpark is open source cuts in its favor: the code can be tested independently, which is more than can be said for closed serving stacks making similar claims. Still, inference speedups are notoriously workload-dependent. Gains that appear on one batch size, sequence length, or model architecture can shrink dramatically on another, and the report available at publication does not specify the conditions behind the headline number.
The practical test is adoption. The inference-serving field already has entrenched open-source incumbents — frameworks like vLLM and NVIDIA’s TensorRT-LLM ecosystem have large communities and production track records. DSpark’s real-world impact will be measured not by its launch benchmark but by whether major serving operations fold it, or its techniques, into production over the following quarters. DeepSeek’s prior open-source releases were rapidly picked apart and partially absorbed by the community; that is the most likely path here too, even if the framework itself does not displace incumbents wholesale.
Background
DeepSeek emerged from High-Flyer, a Chinese quantitative hedge fund, and stunned the AI industry in January 2025 when its R1 model matched much of the reasoning performance of leading Western systems at a reported fraction of the training cost — an announcement that briefly wiped hundreds of billions of dollars from AI-linked stocks as investors reassessed how much compute frontier AI truly requires. The lab has since maintained a strategy of releasing open-weight models and open-source infrastructure tooling, positioning efficiency as its core identity.
The inference-serving market it is now entering more forcefully has its own history: open-source frameworks such as vLLM (from UC Berkeley researchers) and NVIDIA’s TensorRT-LLM became the workhorses of production LLM serving as the industry’s cost center shifted from training models to running them for hundreds of millions of users. Every meaningful gain in serving efficiency ripples outward into GPU procurement, data center planning, and the unit economics of AI products.
Investor Stanley Druckenmiller has disclosed new equity positions in three publicly traded bitcoin miners — Hut 8, Riot Platforms and Bitdeer — according to a Yahoo Finance report dated June 27, 2026. All three companies have been actively repositioning parts of their energized data center footprints toward artificial intelligence and high-performance computing workloads.
Executive Summary
The disclosure matters less for its dollar size, which the source does not quantify, than for the pattern: a well-known macro investor concentrating on three miners that share a common pivot story. Hut 8, Riot Platforms and Bitdeer each control large blocks of contracted power and operational data center sites — assets that have become scarce in a market where AI training and inference demand is running ahead of grid interconnection queues.
For readers outside finance, a stake disclosure of this kind does not commit the manager to a long-term view, nor does it validate any specific company’s execution. It does, however, mark that a discretionary investor with a long macro track record sees enough upside in the miner-to-AI trade to take exposure to all three names rather than pick a single winner.
Why Miners Are Suddenly AI Real Estate Plays
Bitcoin miners spent the last decade acquiring something the AI industry now urgently needs: interconnected sites with signed power contracts, substations, cooling, and the permits to operate at hundreds of megawatts. Building that stack from scratch in the United States or Canada today typically takes three to seven years, dominated by utility interconnection studies rather than construction. Miners already have the electrons, even if their existing buildings were designed for air-cooled ASIC racks rather than liquid-cooled GPU clusters.
That gap — energized land versus AI-ready halls — is the core of the investment thesis. Retrofitting a mining shed for high-density GPU compute is expensive and technically demanding, but it is faster and cheaper than winning a new interconnection. Investors buying the miner-to-AI story are effectively paying for optionality on power, with bitcoin revenue as a floor while sites are converted or leased.
Three Companies, Three Different Bets
Grouping Hut 8, Riot and Bitdeer together is convenient but glosses over meaningful differences. Hut 8 has publicly pursued a diversified compute strategy that includes managed services and AI-oriented capacity. Riot Platforms has historically emphasized scale in Texas mining, with more recent signals toward HPC hosting. Bitdeer combines self-mining, hosting and its own ASIC design, with sites across multiple jurisdictions.
A basket approach — taking positions in all three rather than one — is consistent with an investor who believes the theme will work but is uncertain which operator will convert power into AI revenue most efficiently. It also spreads exposure across different regulatory regimes, customer mixes, and balance sheets, each of which will matter more than the bitcoin price if AI hosting becomes the primary revenue line.
What A 13F-Style Signal Does and Does Not Mean
Position disclosures by well-known investors routinely move share prices, and coverage of this kind tends to be read as endorsement. It is worth being precise about what such a filing conveys: it is a snapshot of holdings as of a past date, without cost basis, without hedges, and without the manager’s forward intent. A stake can be trimmed or exited before the market ever sees the next disclosure.
For infrastructure buyers evaluating these operators as potential AI capacity providers, the more relevant questions are contractual: what tenants have signed, at what power price, on what term, and with what service-level commitments around uptime and density. Those data points, not fund flows, determine whether a converted mining site is a credible enterprise-grade colocation offering.
Background
Publicly traded bitcoin miners emerged as a distinct equity category after 2017, scaling rapidly through the 2020-2021 crypto cycle by locking in long-term power contracts, often in Texas, the U.S. Midwest, Canada and Scandinavia. The 2024 bitcoin halving compressed mining margins and coincided with an unprecedented surge in AI compute demand, prompting several miners to publicly reposition energized sites toward AI and high-performance computing hosting.
Hut 8, Riot Platforms and Bitdeer are three of the most-watched names in that transition. Institutional investor attention to the group has grown as hyperscalers and AI-native tenants search for sites where power is already contracted, since new utility interconnections in North America can take years to secure.
Microsoft is claiming water positivity across its data center operations, according to a June 27, 2026 report from Data Center Dynamics. Water positivity means an operator replenishes more water to stressed watersheds than its facilities consume — a milestone Microsoft first committed to reaching by 2030 when it announced its water-positive pledge in 2020.
The claim spans one of the world’s largest cloud footprints, and it arrives at a moment when AI-driven capacity growth has put data center water consumption under intense public and regulatory scrutiny. The available report is headline-level, so the scope, accounting method, and verification behind the claim remain to be detailed.
Executive Summary
Microsoft has publicly positioned its data center operations as water positive — consuming less water, net of replenishment projects, than it returns to the watersheds where it operates. If the claim holds up under scrutiny, it would represent the first time a hyperscale cloud operator has asserted that its fleet, as a whole, has crossed that line, and it would land years ahead of the company’s stated 2030 target.
Why it matters: water has become the second front, after power, in the fight over data center siting. Communities from Arizona to the Netherlands have pushed back on facilities that draw millions of gallons for evaporative cooling, and regulators increasingly ask for water commitments alongside grid commitments. A credible water-positive benchmark from the market’s second-largest cloud provider would reset expectations for every operator negotiating a site — including colocation and wholesale providers who compete for the same land, power, and permits.
The operative word is credible. Water positivity is an accounting construct, not a physical description of any single site, and its value depends entirely on scope, measurement, and where the replenishment actually happens. The source reporting available at publication does not yet answer those questions, and they are the right ones to ask of any operator making a similar claim.
What “Water Positive” Actually Means — and What It Doesn’t
Water positivity is a ledger claim: over a defined period, the volume of water an operator restores — through wetland restoration, leak-repair programs, irrigation efficiency projects, aquifer recharge, and similar investments — exceeds the volume its operations consume. Consumption here typically means water evaporated or otherwise not returned to the source, which for data centers is dominated by evaporative cooling, the technique of cooling air or water by letting some of it evaporate, trading water for large electricity savings.
What the construct does not mean is that any individual data center stopped drawing water. A facility in a drought-stressed basin can keep consuming while the corporate ledger balances with a restoration project elsewhere. That is not inherently bad-faith accounting — carbon markets work on a similar logic — but water is far more local than carbon. A gallon replenished in one river basin does nothing for the aquifer under a different one. The strongest version of a water-positive claim is basin-matched: replenishment in the same watersheds where consumption happens, weighted toward the most stressed ones. Whether Microsoft’s claim is basin-matched is exactly the kind of detail the headline-level reporting leaves open, and it is the difference between a milestone and a marketing line.
The Cooling Economics Behind the Claim
Data centers face a three-way trade among water, energy, and capital. Evaporative cooling is cheap and energy-efficient but water-hungry. Closed-loop and air-cooled designs eliminate most on-site water consumption but raise electricity use or capital cost, and in hot climates they can strain the power budget that operators are already fighting to secure. Microsoft has spent several years publicizing designs that move toward zero-water cooling for new builds, alongside efficiency metrics like WUE — water usage effectiveness, the liters of water consumed per kilowatt-hour of IT load.
A fleet-level water-positive result, if achieved early, most plausibly reflects three levers working together: newer builds consuming less per megawatt, replenishment portfolios scaling faster than consumption, and — the uncomfortable variable — how fast AI capacity growth adds consumption to the denominator. The AI buildout cuts both ways here. High-density AI halls increasingly use direct liquid cooling, which circulates coolant in a closed loop and can actually reduce on-site water consumption per unit of compute, but the sheer volume of new capacity can swamp per-unit gains. Any operator’s water math in 2026 is a race between those two curves.
A Benchmark With Teeth — If the Methodology Is Public
The industry consequence of this claim depends less on Microsoft than on procurement. Enterprise cloud buyers and public-sector tenders already ask for carbon disclosures; a hyperscaler asserting water positivity gives sustainability teams a new line item to demand from every provider. Google and Amazon have announced their own 2030-era water goals, so competitive pressure to demonstrate progress — not just pledge it — will rise. Colocation operators, who often lack the balance sheet for large replenishment portfolios, may feel the squeeze most: their water story is largely their cooling design, not an offsetting ledger.
For communities and regulators, the useful move is to treat the claim as an invitation to standardize. Today there is no universally accepted audit standard for water positivity comparable to the frameworks maturing around carbon. Claims are only comparable across operators if consumption scope (owned versus leased capacity, construction water, upstream power-generation water), replenishment crediting rules, and basin matching are disclosed. An early, well-documented claim from a market leader could seed that standard. A thinly documented one would invite the same greenwashing skepticism that has dogged renewable energy certificates — and would make life harder for operators doing the work rigorously.
Background
Microsoft is one of the world’s largest data center operators, running cloud infrastructure across dozens of countries to serve its Azure, Microsoft 365, and AI businesses. In 2020 the company pledged to become water positive by 2030 as part of a broader sustainability program that also targets carbon-negative operations, and it has since promoted lower-water cooling designs for new facilities alongside a portfolio of watershed replenishment projects.
The claim lands in an industry racing to build AI capacity while facing growing scrutiny over resource consumption. Water has joined electricity as a gating factor for new data center permits, and no common audit standard yet exists for corporate water-positivity claims — which makes the methodology behind any such announcement as consequential as the announcement itself.
Reuters reported on June 27, 2026 that Chevron, the second-largest US oil and gas producer, is looking at more deals to supply electricity to American data centers. The report signals that Chevron intends to expand beyond its previously announced data-center power venture and treat AI-driven electricity demand as an ongoing line of business rather than a one-off experiment.
Executive Summary
According to the Reuters report, Chevron is actively seeking additional opportunities to power US data centers. The company had already staked out a position in this market: in early 2025 it unveiled a venture with investment firm Engine No. 1 and turbine maker GE Vernova to build natural-gas power plants co-located with data centers — so-called behind-the-meter generation that serves a facility directly rather than routing through the public grid — with a stated ambition of up to four gigawatts of capacity. A statement of appetite for “more deals” suggests that pipeline is progressing well enough for Chevron to widen it.
Why it matters: the binding constraint on AI infrastructure has shifted from chips to electricity. Utility interconnection queues in major US markets now stretch years, and hyperscalers and data-center developers are increasingly willing to contract directly with anyone who can deliver firm power on a faster clock. An integrated oil major brings its own fuel supply, engineering capability, and balance sheet to that problem — a combination few pure-play power developers can match.
From Barrels to Electrons: Why Oil Majors Want AI Load
Oil and gas companies have spent the past decade searching for growth businesses that fit their existing skills. Data-center power is unusually well matched: it monetizes natural gas — which Chevron produces in large volumes, particularly in the Permian Basin — through long-term contracts with creditworthy technology counterparties, and it uses project-development muscle the industry already has. Unlike many diversification bets, it does not require the company to learn an unfamiliar trade; it moves gas one step further down the value chain, from selling the fuel to selling the electricity made from it.
For Chevron, the strategic appeal is margin and duration. Spot gas prices are volatile, but a multi-year power contract with a data-center operator converts that volatility into something closer to an annuity. If AI demand projections hold, an oil major that locks in supply relationships now is positioning itself in one of the few large, growing markets for hydrocarbons in the developed world.
Behind-the-Meter Power: The Speed Play
The core product here is speed. Connecting a large new load to the grid in many US regions means joining an interconnection queue and waiting — often three to five years or more — while studies and upgrades grind forward. Behind-the-meter generation sidesteps much of that by building the power plant at the data-center site, dedicated to that customer. For an AI developer racing to energize capacity, shaving years off time-to-power can be worth paying a premium.
The trade-offs are real, though. On-site gas generation ties the facility’s economics to fuel prices and turbine availability, and gas turbines are themselves in short supply, with manufacturers reporting multi-year order backlogs. It also raises questions for local communities and regulators about emissions, water, and whether large loads that bypass the grid still contribute fairly to shared infrastructure costs. None of these is disqualifying, but each is a live negotiation in every deal of this kind.
The Competitive Field Is Crowding Fast
Chevron is not alone in this pivot. Rival Exxon Mobil has discussed plans for gas-fired plants with carbon capture aimed at data centers, and a broad set of players — independent power producers, private-equity-backed developers, nuclear operators, and the utilities themselves — are all courting the same hyperscale customers. The winners will likely be those who can credibly promise firm megawatts on the shortest timeline, which favors companies with secured turbine slots, owned fuel supply, and sites already in hand.
For data-center operators and their tenants, more competition among power suppliers is straightforwardly good news: more options, more negotiating leverage, and a wider menu of structures from full behind-the-meter islands to hybrid grid-plus-onsite designs. For utilities, it is more ambiguous — every gigawatt served behind the meter is load growth they do not capture, at a moment when load growth had finally returned to their business case.
Background
Chevron is one of the world’s largest integrated energy companies and the second-largest US oil and gas producer, with major positions in the Permian Basin of Texas and New Mexico. Like other oil majors, it has been searching for growth avenues as transportation-fuel demand matures; powering data centers emerged as a candidate in early 2025, when Chevron announced a venture with Engine No. 1 and GE Vernova to build gas-fired plants co-located with computing facilities.
The backdrop is a step-change in US electricity demand. After roughly two decades of flat consumption, AI training and cloud computing have driven forecasts of sustained load growth, while grid interconnection queues and equipment shortages slow conventional responses. That gap between demand and deliverable supply is the market opening that Chevron — and a growing list of competitors — is moving to fill.
Rep. Alexandria Ocasio-Cortez (D-NY) has introduced the AI Data Center Moratorium Act, legislation that — as its name states — would impose a moratorium, or temporary freeze, on new AI data center construction in the United States. The bill was reported by Broadband Breakfast on June 27, 2026.
It represents the most direct federal legislative challenge yet to the AI infrastructure boom, moving opposition from county zoning boards and state utility commissions to the floor of Congress.
Executive Summary
Until now, resistance to AI data center construction has been overwhelmingly local: rezoning denials, water-use disputes, and rate cases before state utility commissions. The AI Data Center Moratorium Act changes the venue. By proposing a federal pause on new builds, the bill converts a patchwork of site-by-site fights into a single national policy question about whether the AI buildout should continue at its current pace.
The bill’s practical odds are a separate matter from its significance. Legislation introduced by a House member in the minority of a policy debate this contested rarely becomes law quickly, and nothing in the initial report indicates committee support or a Senate companion. But introduced bills do three things regardless of passage: they give opposition a national organizing document, they force industry to argue its case in federal terms, and they establish a marker that future Congresses can pick up if public sentiment shifts.
For data center developers, hyperscalers, and the utilities planning decades of capacity around AI demand, the substance of the moratorium matters less right now than the signal: the political cost of the buildout is rising, and it has reached Washington.
From Zoning Boards to Capitol Hill
The AI infrastructure boom has drawn scrutiny wherever it lands — over electricity demand, water consumption for cooling, land use, and the question of who pays for the grid upgrades large facilities require. What has been missing is a federal focal point. Local opposition wins or loses one site at a time; a federal moratorium bill, even one unlikely to pass, nationalizes the argument.
That shift matters because the industry’s siting strategy has partly relied on jurisdictional flexibility: if one county says no, a neighboring one courting tax revenue may say yes. A federal freeze would remove that option entirely, which is precisely why the industry will take the bill seriously as a signal even while discounting it as law. It also invites a counter-response — federal legislators favorable to the buildout may now push preemption or permitting-acceleration measures, making Congress a two-way battleground rather than a bystander.
The Economics a Moratorium Would Collide With
AI data centers sit at the center of enormous committed capital. Hyperscale cloud providers and AI developers have publicly planned multi-year construction programs, and utilities in several regions have built their load forecasts — and their generation and transmission investment plans — around expected data center demand. A construction freeze, if enacted, would ripple through all of it: land already optioned, power purchase agreements already signed, chip and electrical-equipment orders already placed.
Supporters of a pause would frame that as the point — that commitments are being locked in faster than communities and grids can evaluate them, and that a freeze creates space to assess electricity price impacts and resource use before the buildout becomes irreversible. Opponents would argue a moratorium simply exports construction, jobs, and AI capability to other countries without pausing global demand. Both arguments deserve scrutiny against evidence: what a moratorium would actually change depends on details — scope, duration, exemptions — that the initial report does not provide.
What Each Side Still Has to Prove
The bill’s proponents carry a burden of evidence: demonstrating that data center growth is materially raising household electricity rates or straining water supplies in ways existing state and local review cannot manage, and that a blanket federal freeze is a proportionate remedy rather than a blunt one. Grid-cost allocation is genuinely contested territory — some utilities and regulators have moved to special tariffs that make large loads pay their own way, which weakens the case that a moratorium is the only protective tool available.
The industry carries a symmetrical burden. Claims that data centers are net community benefits rest on tax revenue and construction employment, but permanent job counts at data centers are modest relative to their footprint, and confidential agreements around power pricing and incentives make independent verification difficult. If developers want to defeat moratorium politics, the most effective rebuttal is transparency: publishable data on rate impacts, water use, and cost allocation. Neither side’s talking points should be accepted by label alone.
Background
The AI boom that followed the emergence of large language models set off the fastest data center construction wave in the industry’s history, with hyperscale cloud providers and AI developers committing capital on a multi-year horizon and utilities re-planning generation and transmission around expected demand. As facilities grew from tens to hundreds of megawatts — a single large campus can draw as much power as a mid-sized city — friction with host communities grew with them, producing zoning fights, water disputes, and rate cases across the country.
Rep. Ocasio-Cortez has long been associated with legislation linking energy, climate, and economic policy, most prominently the Green New Deal framework. The AI Data Center Moratorium Act extends that posture to AI infrastructure, and marks the first time the buildout’s opponents have consolidated their case into a proposed nationwide freeze rather than site-by-site resistance.
A US government standards agency has published a guide telling water and sewer utilities that their riskiest weak spot is remote access: the ways staff and outside contractors log in from elsewhere to run pumps and treatment equipment.
More than half of American water systems serve 500 people or fewer and have few staff to defend them.
This matters beyond your tap. Many of the giant computer buildings behind artificial intelligence use water to keep their machines cool, often supplied by those same local utilities.
The US National Institute of Standards and Technology (NIST) has published SP 1800-45, Cybersecurity for the Water and Wastewater Sector: Build Architecture, which identifies secure remote access as the immediate cybersecurity priority for water and wastewater utilities, Industrial Cyber reported on June 26. The project, run by NIST’s National Cybersecurity Center of Excellence (NCCoE), offers high-level reference architectures and three example implementations built from commercially available products.
The three designs cover a traditional on-premises gateway (demonstrated with TDI ConsoleWorks), a cloud-managed access model (StrongDM with Cisco Duo), and encrypted machine-to-machine links between remote sites (Q-Net Security hardware). NIST describes them as product-agnostic demonstrations of capability, not vendor endorsements.
Executive Summary
NIST’s new practice guide gives water utilities a concrete starting point for what it calls the most pressing gap in their operational technology (OT), the networks of controllers and sensors that physically run pumps, valves and treatment processes. Utilities have moved from manual, on-site operation to automated and remotely managed systems. That shift improved efficiency but opened two-way paths into equipment that, if misused, could let an attacker manipulate processes or create unsafe conditions.
The guidance matters to the data center industry because water is an upstream dependency for many facilities that use water-based cooling. A data center can harden its own networks and still depend on a municipal control system it neither owns nor sees into. NIST’s focus on remote access marks where that external dependency is most exposed, and its emphasis on staffing and funding signals that many utilities will struggle to close the gap quickly.
The guide is also a snapshot of a market decision facing thousands of utilities: build remote-access security in-house, rent it from a cloud provider, or encrypt the links between sites in hardware. Each choice moves risk to a different place, and at least one vendor is already arguing that none of these designs goes far enough.
Why a Water Security Guide Is a Data Center Story
A water-cooled data center, especially one that relies on evaporative cooling towers, depends on a steady supply of water at usable pressure and quality. That supply is managed by the local utility’s SCADA (supervisory control and data acquisition) system: servers, programmable logic controllers (PLCs, the small industrial computers that switch pumps and valves) and operator screens spread across a main control station, regional centers and remote pump stations. NIST’s point is that these pieces are held together by remote access, both people logging in from outside and machines at separate pumping stations talking to each other over telecom networks. The guide warns that enabling two-way remote access to operational or safety systems widens the attack surface and could allow malicious actors to manipulate processes or create unsafe conditions.
Follow that chain downstream and the conclusion in our headline follows. For a facility whose cooling depends on municipal water, the earliest point at which a cyber incident could reach its cooling is not its own chiller plant. It is the utility’s remote-access layer: the vendor login to an operator screen, the cloud-brokered session, or the unencrypted controller traffic between pump stations. NIST’s three blueprints map directly onto those three entry points, which is why they read as a map of inherited risk for anyone siting or operating water-cooled capacity.
The exposure is uneven. Facilities using air cooling or closed-loop liquid systems draw far less from the municipal supply than those running evaporative towers, and any disruption at a utility would hit hospitals, homes and industry first and just as hard. The practical point is narrower and more useful: for water-dependent sites, the utility’s cyber posture belongs in site selection, resilience planning and supply conversations, alongside drought risk and capacity allocation.
Utility Size Decides Which Blueprint Is Realistic
NIST’s own figures show why no single design fits the sector. Very small community water systems, serving between 25 and 500 people, make up more than half of all systems but serve a small fraction of the population. Fewer than 1 percent of utilities are classified as very large, serving more than 100,000 people, yet they supply nearly half of the US population. Larger utilities tend to run dispersed infrastructure with complex SCADA networks and dedicated IT and OT staff. Smaller ones often rely on simpler or fully manual systems, aging hardware, few staff and tight budgets.
That split drives the choice of architecture. NIST expects larger systems to support in-house deployments, while smaller utilities are likelier to lean on vendor-managed or cloud-based remote access. The consequence is that for a large share of the country’s water systems, security will rest heavily on the terms of a service contract. NIST explicitly warns utilities to check those contracts for encryption, access controls, intrusion detection and vulnerability management. The guide also stresses that adequate staffing and funding are essential to configure and maintain any of these environments, a quiet acknowledgment that the architecture is the easy part.
Three Architectures, Three Places to Put the Risk
The first design is the classic industrial pattern: a demilitarized zone (DMZ, a buffer network between the outside world and the control system) guarded by paired firewalls, with a remote access server inside it. Users authenticate with multi-factor authentication over HTTPS, role-based rules limit which SCADA consoles, operator screens and PLCs they can reach, and every session is logged. NIST demonstrated it with TDI ConsoleWorks. It offers the utility the most control and demands the most in-house skill.
The second moves authentication, authorization and session monitoring to cloud-managed infrastructure. Users sign in with multi-factor authentication, receive an access token, and a security gateway inside the operational network checks that token before allowing a connection. NIST demonstrated this with StrongDM and Cisco Duo. It lowers the management burden, which is why NIST pitches it at smaller utilities, but it shifts part of the trust to the provider.
The third addresses traffic that no person touches: PLCs and remote terminal units exchanging process data between sites. Here NIST recommends encryption appliances at each network edge that authenticate one another and encrypt data before it leaves. The demonstration used Q-Net Security’s Q-Box hardware to protect Modbus traffic between PLCs. Modbus is a long-established industrial protocol that was not designed with built-in encryption, so this design closes a gap that access controls on human users do not address.
The Hardware-Isolation Critique, Weighed Fairly
Mike Carr, field CTO of Xona Systems, agreed that remote access is the water sector’s biggest cybersecurity weakness, but argued that the named products are all software sitting on the same attack surface adversaries already exploit. He called for hardware-enforced protocol isolation that physically separates the user from the critical network, summarised as “You can’t hack what you can’t reach.”
The critique makes a real point: remote-access gateways are themselves internet-facing targets, and adding controls to an exposed path is not the same as removing the path. Several parts of the claim need testing against the guide itself, though. The third implementation is hardware-based. NIST presents all three as product-agnostic examples rather than recommended products. And NIST treats remote diagnostics and vendor maintenance as operational necessities, while noting that some utilities already choose one-way alarming or on-site work for sensitive tasks. The statement that attacks are seen “in the field every day” comes without incident data. Utilities weighing the argument should ask any vendor, including this one, for evidence that its model preserves the maintenance access they actually need.
Background
NIST, part of the US Department of Commerce, publishes widely used cybersecurity frameworks and guidance. Its National Cybersecurity Center of Excellence produces the SP 1800 series of practice guides, which show how commercially available products can be combined to solve specific security problems in a given sector. SP 1800-45 applies that approach to water and wastewater utilities, a critical infrastructure sector made up of thousands of systems that range from tiny rural operators to large metropolitan authorities.
For the data center industry, water has become a central siting and operating issue as capacity grows for AI and cloud workloads. Facilities that cool with water depend on the reliability of local utilities, which puts utility resilience, including the cybersecurity of utility control systems, alongside power availability and land in data center planning.
Louisiana is competing hard for data centers, the giant computer warehouses that run artificial intelligence and use enormous amounts of electricity.
Governor Jeff Landry signed an order saying such projects should lose a state tax break if they push up everyone else’s power bills.
The catch: the biggest bill on the table is a $1.8 billion Texas power plant that the state’s largest utility wants to buy, largely to serve a Meta data center. The order does not touch it. The state’s elected utility regulators decide that one, and they are leaning toward no.
Louisiana Gov. Jeff Landry signed an executive order on June 25, 2026, directing Louisiana Economic Development Secretary Susan Bourgeois to write, within 90 days, new requirements that data centers must meet to claim the state’s Data Center Sales and Use Tax Exemption, the Louisiana Illuminator reported. Under the order, data centers that cause electricity prices to rise across the grid should not qualify for the break unless other benefits outweigh the higher prices.
The order arrives amid a dispute over Entergy Louisiana’s proposed $1.8 billion purchase of the Cottonwood power plant in Newton County, Texas, a price more than three times what the plant sold for in 2024. A consultant to the Louisiana Public Service Commission attributed that increase predominantly to the power needs of Meta’s Hyperion data center in Northeast Louisiana. The order explicitly does not apply to Entergy’s generation plans.
Executive Summary
Landry’s order puts Louisiana in the group of states trying to court hyperscale data centers (the very large campuses built by major cloud and AI companies) while promising that households will not subsidize their electricity. It does this through an economic development lever: a sales tax exemption that projects must now earn by showing they will not raise grid-wide power prices, or that their benefits outweigh any increase.
The order’s reach, however, stops short of the decision that currently matters most to ratepayers. Entergy Louisiana’s plan to buy the Cottonwood plant, which consultant Lane Sisung projected would significantly raise monthly bills for the utility’s 1.1 million customers, is before the Public Service Commission, and the order states it does not limit the commission’s authority. That split between the tax lever and the rate lever is the story’s central point for anyone building, financing or siting AI infrastructure in the state.
The Order Polices a Tax Break, Not a Power Plant
The mechanism in Landry’s order is a condition on a tax incentive. Secretary Bourgeois has 90 days to set requirements for the Data Center Sales and Use Tax Exemption that balance investor interests with residents’ electricity costs. A project that raises prices across the grid is meant to lose eligibility unless its other benefits outweigh that effect. That is a gate for future incentive claims, administered by the state’s economic development agency.
The largest cost item currently in view runs through a different gate. Entergy Louisiana’s proposed $1.8 billion purchase of the Cottonwood plant in Texas, more than triple its 2024 sale price, is a generation decision regulated by the Public Service Commission. Sisung’s analysis for the commission said the price increase is predominantly attributable to power demand from Meta’s Hyperion data center, and projected significant monthly bill increases for Entergy Louisiana’s 1.1 million ratepayers if the deal proceeds. The order says it does not apply to Entergy’s generation plans and does not impair the commission’s constitutional or statutory authority.
So whether those 1.1 million customers absorb costs tied to Hyperion’s load will be settled by five elected commissioners, not by the executive order. Republicans hold a 3-2 majority on the commission and have said they are inclined to follow Sisung’s advice against approving the purchase. For developers and utilities, the practical reading is that Louisiana now has two venues for data center cost questions, and the one with direct control over rates is the commission.
Why an Incentive Condition Is Useful but Indirect
Tying a tax break to grid effects is not empty. Incentives are negotiated before a site is chosen, so a cost test attached to them can shape project terms early, for example by encouraging developers to fund their own generation or accept tariffs that assign new-capacity costs to them. That is leverage a governor has and a commission does not.
The limits are equally clear. Generation needed to serve a large new customer is typically committed through utility agreements and regulatory approvals that the incentive does not govern. Whether a project “causes energy prices to rise across the grid” is also a modeling question, and the order’s balancing test, which allows other benefits to outweigh higher prices, leaves room for judgment. How strict the rule proves to be depends on requirements that have not yet been written.
Every Side’s Claims Deserve the Same Scrutiny
Entergy and Meta have said they disagree with Sisung’s assessment, and Landry has said the Cottonwood proposal would breach promises both companies made not to shift electricity costs to ratepayers. The companies’ disagreement is a claim that needs the same testing as the consultant’s conclusion; the commission’s proceeding is where that evidence should be weighed. Entergy Louisiana CEO Phillip May attended the signing and accepted a copy of the order, while the governor said he had not changed his view on Cottonwood.
The critics’ claims also warrant scrutiny. Commissioner Davante Lewis called the event a “PR stunt” and criticized closed-door data center deals. Jackson Voss of the Alliance for Affordable Energy supported the goal but said Landry pressed the commission last year to shorten its review of Entergy’s power agreement with Meta under the governor’s Lightning Speed Initiative. The Sierra Club called the order “hollow.” The Lightning Speed critique goes to how the current cost question arose; the “hollow” charge is a prediction about rules not yet published, and it can be tested against them once they are.
What It Signals for Hyperscale Siting
For companies evaluating Louisiana, the order adds a cost-allocation test to a financial incentive that matters in site selection. Developers should expect to document grid effects when claiming the exemption, and should treat utility and commission proceedings as the place where their power costs, and who pays them, are ultimately fixed.
The episode also shows the tension facing any state pursuing AI buildout: the order describes data centers as a once-in-a-generation opportunity for rural growth while acknowledging that their power demand can raise costs for existing customers. Louisiana’s answer so far splits that problem between two institutions, and the commission’s Cottonwood decision will be the first real test.
Background
Louisiana has actively recruited large data center projects, most prominently Meta’s Hyperion campus in Northeast Louisiana, which Entergy Louisiana has agreed to supply with power. Gov. Jeff Landry’s Lightning Speed Initiative was created to accelerate permitting for Big Tech and AI investments, and the state offers a Data Center Sales and Use Tax Exemption as an incentive.
Electricity rates for Entergy Louisiana’s 1.1 million customers are set by the Louisiana Public Service Commission, an elected five-member body with constitutional authority over utilities. The commission is now reviewing Entergy’s proposed $1.8 billion Cottonwood plant purchase, the first major test of whether costs tied to data center demand will reach household bills.
WIRED’s Security News This Week roundup for late June 2026 reports that leading AI companies are publicly warning of a cybersecurity ‘apocalypse’ expected within months, tied to the growing capability of AI systems to accelerate offensive cyber operations.
The item appears in WIRED’s weekly security digest dated June 26, 2026, framing the warning as a high-signal alarm from AI vendors themselves rather than from outside researchers or government agencies alone.
Executive Summary
The headline claim is unambiguous: AI ‘giants’ — the large model developers whose systems increasingly power both productivity and, potentially, attack tooling — are telling the public that AI-assisted cyberattacks are about to reach a qualitatively new level, on a timeline measured in months rather than years.
For infrastructure operators, the practical question is not whether AI accelerates certain attacker workflows (it plainly does) but whether the near-term step change is severe enough to justify emergency posture changes. The vendors making the warning are also selling the tools proposed as remedies, which does not make the warning wrong but does mean the evidence should be weighed rather than accepted on authority.
The source we can point to is a single WIRED roundup entry. The underlying vendor statements, threat models, and timelines are not reproduced in the item summary available to us, and readers should treat the WIRED framing as a pointer to a broader conversation rather than a full accounting.
A Warning From Parties on Both Sides of the Trade
When the companies building the most capable AI systems tell the public that those same systems are about to make cyberattacks dramatically worse, the message carries weight — and a built-in conflict. The same firms sell AI-powered defense products, security copilots, and enterprise safety tooling. That does not falsify the warning; capable insiders are often the first to see a problem. But it does mean the claim should be evaluated on the evidence disclosed, not on the identity of the messenger. What specific capabilities have crossed a threshold? Which attacker tasks have been automated end-to-end versus merely sped up? The WIRED entry as we see it is a pointer, not a proof, and the vendor statements it references warrant the same pointed questions any market participant’s alarm would.
What ‘Months’ Would Actually Look Like
Cyber ‘apocalypse’ is a loaded word, so it is worth translating. Concretely, a near-term AI-driven step change would likely show up as: faster and more convincing phishing tailored to individuals; automated discovery and exploitation of known vulnerabilities across large IP ranges; lower-skill operators reaching mid-tier attacker capability; and more effective social engineering against helpdesks and identity workflows. None of these are new categories — they are existing threats with the cost curve bending. For defenders, the meaningful metric is time-to-compromise for a typical enterprise versus time-to-detect and time-to-contain. If attackers compress their side of that equation faster than defenders compress theirs, breach frequency and severity rise even without any single dramatic new exploit.
Implications for Infrastructure and Enterprise Buyers
For data center operators, cloud providers, and connectivity carriers, the operational response to this class of warning is not new tooling so much as accelerated hygiene: enforce phishing-resistant authentication (hardware keys, passkeys) for privileged access, shorten patch windows on internet-facing systems, rehearse identity-provider compromise scenarios, and assume that voice, text, and video pretexting will pass casual sniff tests. Enterprises buying AI security products should ask vendors for measured detection and response improvements against realistic attacker workflows, not marketing demos. The economically rational posture is to treat AI as a general accelerant of both attack and defense, budget accordingly, and avoid both complacency and panic-driven procurement.
The Even-Handed Read
Two things can be true at once. AI genuinely lowers the cost of skilled-looking offensive work, and vendors have commercial reasons to amplify urgency. A ‘months away’ timeline is testable — it either materializes in incident data or it does not — and honest reporting a year from now should revisit it either way. Readers should be wary of two failure modes: dismissing the warning because the messengers benefit from it, and accepting a specific timeline without the underlying threat model. Both errors have costs.
Background
WIRED’s ‘Security News This Week’ is a long-running weekly roundup of notable cybersecurity developments, aimed at both practitioners and general readers. It functions as a curated digest, so its lead items typically point to broader industry conversations rather than exhaustively report a single event.
The backdrop to this particular warning is the rapid rise of frontier AI models since 2023 and the parallel emergence of AI-assisted offensive tooling. By 2026, phishing, reconnaissance, and vulnerability triage have all seen documented uses of generative AI, and the largest model developers have built internal safety and security teams that periodically publish threat assessments. This item sits in that lineage.
Research highlighted by Phys.org on June 26, 2026 warns that rising global temperatures and humidity are undermining one of the data center industry’s most important energy-efficiency strategies: free cooling, the practice of using cool outside air or water to remove server heat instead of running energy-hungry mechanical chillers. As more hours of the year become too hot or too humid for outside air to do the job, facilities worldwide face growing cooling energy demand.
The finding lands at a sensitive moment. Data center construction is accelerating to serve AI workloads, and cooling is typically the largest energy consumer in a facility after the IT equipment itself — so any climate-driven loss of free-cooling hours compounds an already steep power challenge.
Executive Summary
The core claim is straightforward: free cooling only works when the outside environment is cooler and drier than the conditions servers require, and climate change is steadily reducing the number of hours per year when that is true. Heat is only half the story — humidity matters just as much, because evaporative cooling systems, which cool air by evaporating water, lose effectiveness as the air becomes more saturated. Regions that were designed around thousands of free-cooling hours a year are watching that budget shrink.
Why it matters: efficiency assumptions made at design time are baked into a data center for decades. A facility engineered in a climate that no longer exists will either consume more energy than its models promised, lean harder on water, or require retrofit investment. For an industry under scrutiny over electricity and water consumption, the research reframes climate not as a sustainability talking point but as an engineering input — one that belongs in site selection, cooling-system choice, and capacity planning from day one.
Free Cooling Was the Industry’s Efficiency Workhorse
For the past fifteen years, the biggest gains in data center efficiency — reflected in falling PUE, the ratio of total facility power to IT power — came largely from using the outdoors as a heat sink. Air-side economizers pull in filtered outside air; water-side economizers and evaporative systems use cooling towers to shed heat with modest energy input. Hyperscale operators famously sited facilities in cool climates precisely to maximize these hours.
The research reported by Phys.org attacks the durability of that playbook. If the number of hours cool and dry enough for economization declines, chillers run more, and the efficiency gap between a well-sited facility and a poorly sited one narrows in the wrong direction. The gains of the last decade were real, but they were partly a loan from a stable climate — and the terms of that loan are changing.
Humidity Is the Underappreciated Variable
Public discussion of data center cooling fixates on temperature, but wet-bulb temperature — a combined measure of heat and humidity that sets the floor for evaporative cooling — is the more binding constraint. When wet-bulb temperatures rise, evaporative systems must work harder and consume more water for less cooling effect, and in extreme conditions they cannot reach the setpoints servers need at all. That pushes operators back toward mechanical refrigeration exactly when grid demand for air conditioning also peaks.
This has a second-order consequence: the trade-off between energy and water gets sharper. Evaporative cooling saves electricity but consumes water; dry coolers and chillers save water but consume electricity. Rising humidity degrades the attractiveness of the water-based option in many regions, forcing a choice between two increasingly expensive resources — often in communities already contesting data center water use.
Winners: Liquid Cooling, Cool Geographies, and Honest Modeling
If outside air can carry less of the load, the premium shifts to technologies that tolerate warmer heat rejection. Direct-to-chip liquid cooling and immersion cooling move heat in water or fluid rather than air, allowing higher operating temperatures and, in many designs, year-round heat rejection without compressors even in warm climates. The AI build-out was already pushing the industry toward liquid cooling for density reasons; climate trends add an efficiency rationale.
Geography gains value too. Sites in cool, dry, or high-latitude regions — the Nordics, parts of Canada, high-altitude locations — become relatively more attractive, though they bring their own constraints in connectivity, latency, and power availability. And engineering firms that model cooling against forward-looking climate projections rather than historical weather files gain a real advantage: a 25-year asset should be designed for the climate of 2040, not 2010.
Risks: Stranded Efficiency and Rising Operating Costs
The losers in this shift are facilities whose economics depend on free-cooling assumptions that no longer hold — particularly older air-cooled sites in regions warming fastest. Their operating costs drift upward without any change in workload, and their sustainability reporting deteriorates through no operational fault. For colocation providers, whose customers increasingly scrutinize PUE and water metrics in procurement, that drift is a competitive problem, not just an engineering one.
There is also a grid-level risk. The hours when data centers lose free cooling are the same hot hours when regional grids are most stressed. Climate-driven cooling demand is therefore correlated demand — it arrives when power is scarcest and most carbon-intensive, which is precisely the scenario utilities and regulators planning for data center growth need to model.
Background
Data center cooling has evolved through distinct eras. Early facilities ran cold and relied almost entirely on mechanical chillers. From roughly 2010 onward, hyperscale operators drove a revolution in economization — siting in cool climates, using outside air and evaporative systems, and widening acceptable server temperature ranges — which pushed the best facilities’ PUE from around 2.0 toward 1.1. That efficiency story became central to the industry’s answer to critics of its energy footprint.
The current AI build-out is testing every part of that model: rack power densities have jumped severalfold, cooling loads are climbing, and communities are scrutinizing both electricity and water consumption. Research showing that climate change is eroding free cooling adds a structural pressure on top of a cyclical boom — and helps explain the industry’s accelerating shift toward liquid cooling and climate-aware site selection.
Research firm Rystad Energy projects that investment in fuel cells by data-center operators will grow roughly tenfold, reaching $30 billion by 2030, according to a report published June 26, 2026. The forecast points to on-site power generation moving from a niche backup strategy to a mainstream way of energizing new data-center capacity as connections to the electric grid grow slower and harder to secure.
Executive Summary
Rystad Energy, a Norway-based energy research and intelligence firm, has put a headline number on a trend the data-center industry has been living with for several years: when the grid cannot deliver power on the timeline a project needs, operators increasingly buy their own generation. Its new forecast calls for data-center fuel cell investment to grow tenfold by 2030, reaching $30 billion — a figure that implies today’s spending is on the order of a few billion dollars a year.
Fuel cells convert a fuel — most commonly natural gas today, potentially hydrogen in the future — directly into electricity through an electrochemical reaction rather than combustion. That gives them attractive properties for data centers: they can be deployed in modular blocks at the site, run continuously as primary power rather than just backup, and generally face lighter air-permitting burdens than combustion turbines or diesel generators. A tenfold growth call, if it materializes, would make fuel cells one of the fastest-growing categories of behind-the-meter power — generation installed on the customer’s side of the utility connection — in the broader AI-infrastructure buildout.
The Grid Queue Is the Real Story
The most important context for this forecast is not the fuel cell itself but the waiting line in front of it. In many major data-center markets, utilities and grid operators have quoted multi-year waits for large new interconnections — the formal process of hooking a big load up to the transmission system. For an AI data center whose revenue depends on being energized quickly, a delay of several years is often more costly than paying a premium for on-site generation. That inversion of economics — time-to-power mattering more than cost-per-megawatt-hour — is what turns a niche technology into a $30 billion market forecast.
Fuel cells are one of several answers to that problem, alongside gas turbines, reciprocating engines, and eventually small modular nuclear reactors. Their particular appeal is speed and siting flexibility: modular units can be added in increments as a campus grows, they operate quietly with no combustion exhaust plume, and in many jurisdictions they clear environmental permitting faster than combustion alternatives. For operators, that can compress the gap between breaking ground and serving customers.
What Tenfold Growth Would Actually Require
Growing an equipment market tenfold in roughly four years is not just a demand question — it is a manufacturing and supply-chain question. Fuel cell systems depend on specialized components and materials, and stepping up output by an order of magnitude means new factory capacity, expanded supplier networks, and trained installation and service workforces. The release headline does not indicate whether Rystad’s forecast is constrained by manufacturing capacity or is a pure demand-side projection, and that distinction matters a great deal for whether the number is achievable.
The fuel supply side deserves equal scrutiny. Most commercially deployed data-center fuel cells today run on natural gas, which means large deployments need pipeline capacity and gas contracts — their own version of an interconnection queue. Operators are effectively trading one infrastructure dependency for another. That trade often still makes sense, because gas infrastructure can frequently be expanded faster than high-voltage transmission, but it is not a free pass around the physical world.
Winners, Losers, and the Emissions Question
If the forecast is directionally right, the clearest beneficiaries are fuel cell manufacturers and the developers who package on-site generation into ready-to-run power solutions for data centers, along with gas utilities that supply the fuel. Traditional electric utilities face a more nuanced picture: behind-the-meter generation can relieve pressure on constrained grids, but it also diverts what would have been decades of steady load growth — and the revenue that comes with it — away from the regulated system.
The environmental ledger is genuinely mixed and worth stating plainly. Natural gas fuel cells emit carbon dioxide, though generally with higher electrical efficiency and far lower local air pollutants than combustion generation. Advocates point to a future switch to hydrogen as a path to low-carbon operation; skeptics note that low-carbon hydrogen remains scarce and expensive. Buyers and communities evaluating these projects should ask which fuel is actually contracted today, not which fuel is possible in principle.
A Forecast Is a Scenario, Not a Commitment
It is worth being clear about what a research-firm projection is: a modeled scenario built on assumptions about data-center demand, grid-connection timelines, technology costs, and competing options. Rystad is a well-established energy intelligence firm, but the headline figure arrives without published methodology in the source at hand. If AI capacity growth slows, if utilities accelerate interconnections, or if gas turbine supply loosens, the fuel cell number could land well short of $30 billion. Conversely, if grid queues lengthen further, it could prove conservative. The forecast is best read as a signal about the direction and seriousness of the on-site power trend, not as a precise measurement of the future.
Background
Data-center electricity demand has surged with the AI buildout, and in several major markets the ability to get grid power — not land or capital — has become the binding constraint on new capacity. That has pushed operators toward on-site generation of many kinds, from gas turbines to fuel cells, and made “time to power” a core competitive metric. Fuel cells entered the data-center world primarily as clean backup and supplemental power, with a small number of vendors building a commercial track record over the past decade; the shift Rystad describes is their promotion to primary, at-scale power for new facilities.
Rystad Energy, founded in Oslo in 2004, built its reputation on oil and gas market intelligence and has since expanded into power, renewables, and energy-transition research, making it one of the more frequently cited independent forecasters in the energy sector.
